Audit of deep wound infection following hip fracture surgery

D.C. MACKAY, W.J. HARRISON, J.H.J. BATES and D. DICKENSON
Orthopaedic Department, Newcastle General Hospital, Newcastle upon Tyne, U.K.

An audit of 171 consecutive hip fractures treated surgically showed a deep wound infection rate of 3.6%. Several shortcomings in the implementation of the infection policy were identified and recommendations to eliminate them introduced. Review of the subsequent 186 patients gave a deep infection rate of 1.1%. Attention to detail and good practice are essential to maintain low wound infection rates.

Deep or major wound infection is a well recognised complication of the surgical treatment of femoral neck fractures. Individual cases place a high burden on hospital resources and result in significant patient morbidity and mortality. However, advances in theatre design and practice, surgical techniques, antibiotic prophylaxis and wound care have served to reduce the reported incidence in recent series to between 0.8 - 1.7%.^1-5 The resultant rarity of individual cases can easily lead to complacency. Constant vigilance and strict adherence to infection prevention policy is necessary to avoid lapses in these high standards of practice. A cluster of deep wound infections among patients undergoing surgery for hip fractures raised concerns that there might be a problem with the infection control policy in this unit. Consequently, a detailed audit was undertaken to establish the true deep wound infection rate, to identify any shortcomings in current practice, recommend ways of eliminating these and finally to re-audit and quantify any improvement.

A retrospective review of the case notes of all patients under-going surgery for proximal femoral fractures between February and August 1997 was undertaken three months after surgery. Details of age, sex, patient risk factors, time from admission to operation, surgical procedure, grade of principle surgeon, use of prophylactic antibiotics and drains, length of stay and mortality rates were recorded.

Cases of deep wound infection were identified clinically. The definition of a “deep” or “major” wound infection proposed by the Surgical Infection Study Group⁶ was used: “a purulent discharge complicated by spreading cellulitis, systemic upset or wound breakdown”. All but one of the cases identified required surgical debridement of the wound, allowing direct bacteriological confirmation of deep infection. The remaining patient suffered a partial wound dehiscence with proven growth of S. faecalis.

Cases of minor wound problems, including local erythema or serous discharge in the early post-operative period, which resolved without treatment, were excluded from the definition of deep wound infection.

It was felt that most infections would have declared themselves by three months. As there was a significant mortality rate prior to three months and a proportion of patients failed to attend for review, the patient’s general practitioner was contacted and the medical records scrutinised, to ensure that all infections were identified. In these cases, details of the organism involved and subsequent treatment were recorded. Initially, it was also intended to audit superficial wound infection rates. However, a precise definition to allow identification of these from the case notes proved difficult and this aspect was not covered.

Analysis of this initial patient group showed a worryingly high deep wound infection rate of 3.6%.

The existing unit policy for prevention of infection included guidelines on the treatment of concurrent infections and optimising the patient’s condition prior to surgery; undertaking all surgery in a laminar flow theatre; strict theatre discipline and sterile technique including the use of waterproof disposable drapes and the wearing of face masks; prophylactic antibiotics; and post-operative wound care.

Scrutiny of these and the actual practice identified several potential problems. The laminar flow was often disrupted by the anaesthetic team being positioned inside the hood area, excessive movement in theatre and the use of the main theatre doors during cases. Instruments were occasionally laid up in the main theatre during transfer and positioning of the patient. Surgeons and scrub staff often wore theatre caps allowing exposure of ears and hair. Some surgeons only single gloved. Poor skin preparation technique with moving from the periphery back to the centre of the wound was identified.

Prophylactic antibiotics were occasionally not given at induction or post-operatively. Wounds tended to be exposed excessively during the early post-operative days for unnecessary wound inspections.

Following discussion between the surgeons, theatre and ward staff revised guidelines designed to eliminate these shortcomings were implemented as shown in Table 1.

Theatre layout	Patient’s head and anaesthetist outside laminar flow hood Main door closed during operations Minimise movement in and out of theatre
Instruments	Instruments covered until patient positioned on table
Clothing	Full hoods and masks for all personnel inside laminar flow
Antibiotics	Cefuroxime 1.5g at induction and 2 post-operative doses
Gowning	Disposable gowns Closed double gloving - 2nd pair after preparing & draping
Skin prep	2 sponges beginning at wound and working out
Draping	Vertical isolation drape for DHS / cannulated screws Disposable hip drapes for Thompsons
Wound care	Dressings should be left closed on ward for 5 days

A subsequent audit of patients between December 1997 and May 1998 was made to assess the effectiveness of the new policy. Differences between the groups were tested for significance using Fisher’s exact probability test and Student t test, as appropriate.

A total of 171 patients were studied in the initial audit group and 186 following the implementation of the new guidelines. Both groups were comparable with no significant differences in patient age, sex, risk factors for infection, interval from admission to operation, overall length of stay and mortality rates at three months (Table 2).

Table 2: Patient factors before and after the introduction of revised guidelines

	Before	After	Significance
Number	171	186
Mean age	76yrs	80yrs	NS
Sex(F/M)	82%/18%	83%/17%	NS
Risk factors
Dementia	16%	20%	NS
Concurrent infection	6%	9%
Diabetes	5%	6%
Steroids	2%	1%
Skin ulcers	1%	1%
Other	5%	7%
Mean time to surgery	1.6 days	1.5 days	NS
Mean length of stay	22 days	19 days NS	NS
Mortality at 3 months	10%	9%	NS

Deep wound infections occurred in 6 patients (3.6%) from the initial audit group as detailed in Table 3. Four of the patients had risk factors for infection. Other than increased age, no specific risk factors could be identified in the other two cases. One patient waited four days prior to surgery, the others being treated swiftly. Two patients failed to receive prophylactic antibiotics. All cases had wound drains inserted. The grade of surgeon involved in the infected cases was no different from the overall distribution of grade of surgeon. The same surgeon was involved in two of the cases. However, scrutiny of the infected cases did not identify a link with a specific individual or strain of organism.

Table 3: Details of patients with deep wound infections before and after introduction of revised guidelines

	Before revised guidelines						After revised guidelines
Case number	1	2	3	4	5	6	1	2
Age	84	86	77	71	94	86	84	83
Sex	F	M	F	F	F	M	F	F
Risk factors	Crohns	Dementia	Steroids	Nil	Nil	Osteomyelitis distal femur 2yrs previously	Nil	Dementia Urinary tract infection
Interval to operation	2	1	2	4	0	1	1	2
Operation type	DHS	Thompson	Thompson	Thompson	DHS	DHS	Thompson	DHS
Grade of surgeon	Reg	Reg	Reg		Reg		Reg	Reg
Antibiotic prophylaxis	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Wound drain	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Bacteriology	S Aureus	S Faecalis	S Faecalis E Coli	S Faecalis E Coli	S Aureus	S Aureus Pseudomonas	S Aureus	S Faecalis Coliforms
Therapy	Debridement	Antibiotics	Girdlestone	Debridement	Girdlestone	Girdlestone	Girdlestone	Debridement
Length of stay (days)	42	54	95	75	78	135	78	83

Re-audit following strict implementation of the new guidelines showed a deep wound infection in only two patients (1.1%). One of these was demented and had a urinary tract infection. The other showed no risk factors. Operation was undertaken early. Both patients received antibiotic prophylaxis and had wound drains inserted. The operations were performed by different surgeons and theatre staff. Although marked, this reduction in the deep infection rate failed to reach statistical significance (p=0.07). The surgical procedure undertaken and the grade of principal surgeon showed no overall differences. The administration of prophylactic antibiotics improved from 91% to 98% and the use of drains declined from 84% to 72% (Table 4).

Table 4: Surgical factors before and after the introduction of revised guidelines

	Before	After	Significance
Operation
Cannulated screws	24%	19%	NS
DHS	40%	49%
Thompsons	34%	30%
Charnley	1.4%	1.5%
Girdlestone	0.6%	0.0%
Grade of surgeon
Consultant	9%	6%	NS
Sp Reg	67%	78%
SHO	24%	16%
Antibiotics given	91%	98%	p<0.05
Wound drain used	84%	72%	p<0.05

Low deep wound infection rates following hip fracture surgery can be achieved by the routine use of a variety of preventative measures.^1-5 This can be achieved despite the predominance of an elderly population with a high incidence of pre-existing medical problems and identifiable risk factors for infection. The final 6 month audit of patients in our unit showed that low infection rates can be attained.

This improvement from the initial audit appears to be independent of any differences in the patient population. The age, sex, interval from admission to surgery, length of stay and mortality rates were comparable in both groups. Pre-existing risk factors for infection were similar, being if anything slightly greater in the final audit group. The surgical procedure undertaken and the grade of principal surgeon were also comparable. The recorded reduction in wound infection rates, therefore, is attributed to the introduction and rigorous enforcement of the revised guidelines. Although not perfect, compliance with antibiotic prophylaxis was improved achieving 98% compliance. The reduction in the use of wound drains was unexpected as their use was not dictated in the policy, but left to the discretion of the surgeon. The decline may indicate a change in practice and could be of relevance. Although difficult to quantify, the guidelines designed to eliminate lapses in theatre practice were rigorously applied and should have served to minimise the risk of intra-operative wound contamination. The policy of not uncovering wounds for 5 days except to replace blood saturated dressings (and then under strict sterile conditions) probably helped to prevent post-operative contamination.

As these changes were made simultaneously it is not possible to ascertain the relative importance of the individual factors. However, the combined effect was to produce a marked decrease in our wound infection rate.

No previous report has set out to study infection post hip fracture surgery as a principal outcome measure. It is hoped, therefore, that our experience and data will provide a useful guide for other units to critically appraise their practice and results.

In conclusion, prophylactic measures can reduce wound infection rates to low levels. However, without strict adherence to these principles the cumulative effect of minor lapses can negate these advances. Good practice minimises wound infections.

Correspondence: Mr D Mackay, 21 Lindale Avenue, Whickham, Newcastle upon Tyne, NE6 5QT, UK

�2000 The Royal College of Surgeons of Edinburgh, J.R.Coll.Surg.Edinb.,45; 1: 56-9

AUDIT

Audit of deep wound infection following hip fracture surgery